A piece I’ve adored for many years. I always thought it could sound great with just a continuo organ and so simply played, pretty much verbatim, the orchestral score from “Georg Friedrich Händels Werke” Band 46a. Bars 28–end tripped me up quite a few times, especially considering how involved the bass gets.

Recording the organ part first presented an interesting challenge: how to keep tempo on what is basically just a drone? A few attempts with a metronome produced only robotic renditions — despite its simplicity, this is a piece with myriad opportunities for tasteful rubato and, of course, ornamentation — and so I ended up recording a scaffolding melody based on Iestyn Davies and Alison Balsom’s recording.

I did miss the syncopation in the bass in beats 1 and 2 of bar 37, but as far as I can hear, the Davies–Balsom recording also doesn’t do it. I wonder what the manuscript says; it is held in the British Library under reference R.M.20.g.2. It seems they did digitise it but it may be unavailable due to the cyber-attack of 2023.

Organ recorded on 2025-12-29. Rather gravelly voice recorded in 3 takes on a rather sick evening of 2026-01-06: likely the last day before chest infection takes my voice, as I can feel it. I was quite seriously debating whether to even attempt recording the voice part, but something in me really wanted to try, and I am at least moderately satisfied with how the intonation turned out.

The trills are of course rather disappointing and this is something I’ve never been able to figure out with my voice.

As of 2026-01-06 I am still undecided whether to attempt recording the trumpet part on trumpet, which would most likely turn out unsuccessful, or something else. I am quite sure I will need to layer on a string ensemble (2 tracks of each?) over the organ, though: organ flute stops alone are simply not enough, texturally. Thankfully, the only viola note that can’t be played on a violin is the F# in bar 28 and that’s easily worked around.

Registration

Sampleset: “Friesach” by Piotr Grabowski.

• Pedal:
  • Gedackt 8'
• Manual II:
  • Nacht. Gedackt 8'
• Couplers:
  • II–P

Recording sessions

As of 2026-01-06 I was undecided whether to attempt recording the trumpet part on trumpet, which would most likely be unsuccessful, or something else. Ultimately the flute seemed a fair choice: if normally a high voice is paired with a relatively mellow, but textured, sound of trumpet, then perhaps it’s reasonable to pair a lower, more textured voice with the purer sound of the flute.

I had originally intended only to have the organ playing the accompaniment, but quickly decided that the flute stops alone were not quite enough, texturally — particularly with a flute also playing the solo part. Thankfully, the only viola note that can’t be played on a violin is the F# in bar 28 and that’s easily worked around.

• Organ
  • Recorded on 2025-12-29.
• Voice
  • Rather gravelly voice recorded in 3 takes on a rather sick evening of 2026-01-06: likely the last day before chest infection takes my voice, as I can feel it. I was quite seriously debating whether to even attempt recording the voice part, but something in me really wanted to try, and I am at least moderately satisfied with how the intonation turned out. The trills are of course rather disappointing and this is something I’ve never been able to figure out with my voice.
  • Re-recorded (to slightly better effect) on 2026-01-16. The breaks in my passaggio are distractingly evident, though; I was still coughing a bit.
  • Recorded again, in rather better voice, on the afternoon of 2026-01-23. Take 26, which I consider acceptable, was actually recorded immediately after getting home from a very tiring weights workout session; after recording that one take, I actually started falling asleep.
  • I persisted in recording on 2026-01-26 despite a dry throat, to little success. The decline in quality vs. 2026-01-23 is stark and really shows that, despite my eagerness to complete the project, it is ultimately a waste of time to try recording when not feeling optimal.
  • I finally found acceptability in take 51, recorded on the morning of 2026-01-27. A few weak notes still caused by a persistently dry throat — this is surely due to me being back in the polluted air of Hong Kong rather than any remaining illness.
  • I decided to try my luck again on 2026-01-28 and take 55 ended up being the final, despite a few notes (particularly in the opening phrase) being a bit on the flat side (though surprisingly this does mesh quite well with the organ, which is also just a bit on the dark side). I am pleased I managed to achieve the octave grace note in bar 13.
• Strings
  • Recorded, to mediocre quality (due to questionable intonation), on 2026-01-15. The strings being just 10 cents sharp turns out to be a bit of a deal-breaker when there are so many open-string notes.
  • Re-recorded on 2026-01-20 in the evening and 2026-01-21 in the morning.
• Flute
  • Recorded 2026-01-19 to minimal satisfaction.
  • Recorded properly on 2026-01-20 in the morning.

Editing

Mixing this has been interesting; the crucial observation is that almost every professional recording I’ve listened to emphasises the basso continuo over the higher strings, and places the voice and trumpet right at the front of the stage. It was quite a shock to me to see how quiet I could make the high strings whilst still maintaining their presence and that veil of sound: judging visually by the sliders and meters alone, I would have thought they’d be inaudible at -24 dB. Yet they absolutely are not.

This is also one of the few pieces I know of that works really, really well with maximum reverb.

Reflections

This is easily the most challenging recording I’ve ever done. It is bizarre that what looks so simple on paper — 38 bars, no flashy virtuosity, only a few instruments to match with, a total vocal range of only a 9th(!)… — could be so difficult to execute properly. Yet it is precisely that simplicity — that sparseness of instrumentation — that leaves every part exposed. It’s much the same reason why some soloists would rather play Paganini than Mozart: when you have a thousand notes and a full Romantic orchestra, one slipped note easily goes unnoticed.

Firstly, the tuning is completely alien to me: this is written for an incredibly sparse orchestra and full of perfect 5ths, which leave very little room for error; there is also the tension between the unjust 5ths of a keyboard instrument and the desire for just 5ths on the strings. Having the strings and organ play identical notes makes, in any case, for a very difficult time on the strings, because things like the sharpened leading note, compressed 3rds, etc. need to be avoided; yet this contradicts the years of practice I’ve had in orchestral and solo settings.

There is also the fact that the 2 solo parts are not really solo: they are firstly a duet with each other, which requires again far more precise intonation than true soloistic playing (which has far more leeway for expressive intonation — a freedom I’m very used to having, since my starting point in classical singing was Romantic-era opera and lieder); but they are secondly also completions of the chords for which the organ and orchestra are the foundation, again demanding effort in producing pure intervals.

Secondly, the voice part represents some of the slowest coloratura I’ve ever attempted. J. S. Bach’s “Ich habe genug” contained, for the longest time, the most difficult passages I knew in terms of quantity of notes and breath control, but those are fast runs that, unlike this piece, don’t have multiple “checkpoint” beats where the voice has to come in perfectly in tune with every other instrument (e.g. second beat of bar 10; third beat of bar 10 in the voice, which has to match perfectly with the flute’s first beat of bar 11…).

Breath control is of course a challenge, and I never thought, in arranging this piece, that I’d be able to sing bars 9–13 in a single breath; yet I surprised myself by making it happen without fail on every single take.

By far more difficult for me was the fact that so many phrases are right in my passaggio (crucially bars 17–27), which means a high risk of cracking notes and unstable intonation.